The relative value of environmental context reinstatement in free recall (El valor relativo de la reinstauración del contexto ambiental en el recuerdo libre)

Ángel Fernández

Description

Abstract
The effect of environmental context on episodic memory was examined in two free recall experiments with groups of old and young subjects. All subjects studied a list of unrelated words and were subsequently tested in the
same room or in a different room. The results of Experiment 1 showed an advantage of being tested in the same context for the old subjects, but no effects of the context manipulation for the young subjects. Experiment 2
replicated this finding and additionally showed that old subjects (but not young subjects) benefited from instructions to mentally reinstate the learning context. The results of both experiments are discussed in terms of
the relative value of contextual cues for subjects in each of the two age groups.
Resumen
Se investigó el efecto del contexto ambiental en dos experimentos de recuerdo libre con grupos de sujetos jóvenes y viejos. Todos los sujetos estudiaron una lista de palabras no relacionadas y posteriormente hicieron pruebas de recuerdo libre en la misma habitación o en una habitación diferente. Los resultados del Experimento 1 demostraron la ventaja de hacer el test en el mismo contexto para el grupo de sujetos viejos, pero no se encontraron efectos de la manipulación contextual con los sujetos jóvenes. En el Experimento 2 se replicó este hallazgo y, además, se demostró que los sujetos viejos (pero no los jóvenes) se beneficiaban de instrucciones de reinstauración mental del contexto de aprendizaje. Los resultados de ambos experimentos se discuten en términos del valor relativo de los indicios contextuales para los sujetos de cada uno de los dos grupos de edad.

Sujets

• Free recall
• Ageing
• Senescencia

Informations

• Publié par: erevistas
• Publié le: 01 janvier 2001
• Nombre de lectures: 10
• Langue: English

Extrait

Psicológica (2001), 22,253-266.
The relative value of environmental context
reinstatement in free recall
1 2Ángel Fernández y María A. Alonso
1 2 Universidad de Salamanca Universidad de La Laguna
The effect of environmental context on episodic memory was examined in
two free recall experiments with groups of old and young subjects. All
subjects studied a list of unrelated words and were subsequently tested in the
same room or in a different room. The results of Experiment 1 showed an
advantage of being tested in the same context for the old subjects, but no
effects of the context manipulation for the young subjects. Experiment 2
replicated this finding and additionally showed that old subjects (but not
young subjects) benefited from instructions to mentally reinstate the
learning context. The results of both experiments are discussed in terms of
the relative value of contextual cues for subjects in each of the two age
groups.
Key words: Environmental context, free recall, aging.
The incidental context in which information is acquired is usually
associated to the learnt material, with the result that reinstatement of the
original learning context can have a favorable effect on later retrieval attempts.
The facilitatory effect of incidental context on the recall of verbal material was
originally supported by the results of a number of classic experiments using
environmental context manipulations of different sorts (for example, Godden
& Baddeley, 1975; Smith, Glenberg & Bjork, 1978).
However, some subsequent studies using similar manipulations have
failed to replicate earlier results. For example, Fernandez and Glenberg (1985)
could not find contextual effects in recall in a lengthy series of experiments in
which environmental context was manipulated by either changing rooms
between study and test or keeping them constant. No reliable differences were
found between subjects who learnt and recalled words in the same room and
subjects who learnt in one room and were tested in a different one. This
absence of contextual effects in situations in which they were expected to

1 Author contact: Angel Fernandez. This research was supported by the Spanish Ministry of
Education and Culture, Grant PB96-1279, and by Junta de Castilla y León, Grant
SA36/00B. We thank anonymous reviewers for their helpful comments and suggestions
on earlier versions of this article. Correspondence should be sent to Angel Fernandez,
Facultad de Psicologia, Universidad de Salamanca, E-37071 Salamanca, Spain (e-mail:
[e-mail]).254 A. Fernández y M.A. Alonso
occur have also been reported in a number of other studies using the same
type of context manipulations and the same type of learning material (for
example, Alonso & Fernández, 1997; Bjork & Richardson-Klavehn, 1989;
Fernández & Alonso, 1994).
The disparity in the data is difficult to explain at the present time, but
some progress could be made along the lines of recently proposed theoretical
accounts. Of particular interest is the “outshining hypothesis” (Smith 1988,
1994), according to which contextual retrieval cues have a relative value and
are most effective when subjects have no better cues available. On the basis of
this idea, it could be argued that, on certain occasions, subjects tested in a
same context condition are not better than subjects tested in a different context
condition simply because neither group makes a significant use of contextual
cues at retrieval. Instead, subjects from both groups may relay more on
alternative cues, either externally provided or internally generated. For
example, the very nature of the memory test can render contextual cues
irrelevant by providing subjects with particularly efficient non-contextual cues,
such as copy cues in recognition tasks or experimentally associated cues in
cued recall tasks. And even in free recall tests, much impoverished in terms of
the external cues they explicitly provide, subjects have the possibility of
applying strategies or retrieval plans that, without making use of contextual
cues, could allow them to meet the demands of the memory task to a
reasonable level.
The outshining hypothesis has been used as a broad interpretive
framework in trying to account for the fact that environmental context effects
are rarely observed in recognition and cued recall tests, and to justify why
many times those effects have proved difficult to find in free recall situations
(Smith, 1988). It has also been used to derive predictions about the effect of
context on the free recall of verbal material previously learnt under different
encoding conditions. For example, McDaniel, Anderson, Einstein and
O’Halloran (1989) manipulated encoding conditions in a series of 5
experiments, hypothesizing that operations that afforded the
generation of richer internal cues (through the formation of visual images,
organizing, self-referencing, etc.) would make recall less dependent on
external context manipulations. They found some support for the predictions,
but, because significant context effects were infrequent and also inconsistent,
it is difficult to arrive at definitive conclusions regarding the adequacy of the
outshining hypothesis on the basis of the results of this study.
The results of another attempt to empirically test the outshining
hypothesis have been more recently reported by Cousins and Hanley (1996).
In their study, subjects studied lists of words under two different encoding
conditions. In the item-processing condition subjects rated individual words
for pleasantness. In the relational-processing condition subjects sorted the
words into pre-specified categories. After a short retention interval, subjects
were given an unexpected free recall test either in the learning room or in a
different one. It was expected that subjects in the relational-processing
condition would be able to use a category-based retrieval strategy and would
not be likely to be affected by the presence or absence of contextual cues atEnvironmental context and recall 255
recall. On the other hand, subjects in the item-processing condition, lacking
the opportunity of using the categorical retrieval strategy, were expected to be
more dependent on the availability of contextual cues and, therefore, more
likely to be affected by contextual manipulations. The data from two similar
experiments showed that level of recall was higher following relational
processing, but not affected by contextual changes in any of the experimental
conditions. These results led Cousins and Hanley (1996) to conclude that the
outshining hypothesis cannot offer an adequate account of the mechanisms
that underlie environmental context effects, at least in free recall situations.
However, it is possible that the procedure used by Cousins and Hanley (1996)
does not allow for a clear test of the predictions derived from the outshining
hypothesis. Their data certainly suggest that relational-processing instructions
led to the use of a category-based retrieval strategy, as evidenced by higher
recall scores and higher clustering scores. But there is no direct evidence
regarding the strategies that could have been used by the subjects in the
itemprocessing condition. The fact that one concrete internally-generated strategy
was not available does not imply that other, non-contextual, internal strategies
were not used by those subjects. Actually, both the type of material used
(imageable nouns) and the type of processing required (pleasantness ratings)
are likely to allow for the generation of rich internal cues at encoding that
could be of use at retrieval (cf. McDaniel et al., 1989). Those cues can be less
efficient than categorical cues in terms of the amount of recall they permit, but
they can be equally effective in making contextual cues less relevant.
A more direct way of controlling the utilization of internally-generated
retrieval cues in environmental context experiments could be the use of groups
of subjects that are particularly impaired in the use of self-generated retrieval
cues. In terms of the outshining hypothesis, this type of subjects should show
a tendency to compensate for the lack of internal cues by relaying more on
contextual cues, and would therefore be more likely to show reliable effects of
contextual manipulations. A review of the literature suggests that the elderly
could be an interesting population in regard to the assessment of this idea. A
consistent finding in studies of memory and aging is that older subjects show
a poorer memory performance than young subjects in a variety of episodic
memory tasks (Poon, 1985), and this difference between the two groups can
be, in part, related to the fact that the elderly are more dependent on external
retrieval support than are younger adults (Craik, 1986, 1992). Consequently,
manipulations such as environmental context changes between study and test,
which can have a direct effect on the availability of external cues, should have
a noticeable impact on the memory performance of older subjects, particularly
in free recall tests, when no much additional retrieval support is provided.
This prediction has been empirically tested in two studies. Phillips and
Kausler (1992) had youn